Kvantatsioonihäire

Kvantatsioonihäire, also known as quantization error, is a phenomenon that occurs in digital signal processing when a continuous analog signal is converted into a discrete digital signal. This conversion process, called analog-to-digital conversion (ADC), involves sampling the analog signal at regular intervals and then assigning a discrete numerical value to each sample. Since the digital system can only represent a finite number of values, the original analog value must be rounded or truncated to the nearest representable digital value. This rounding or truncation process introduces an error, which is the quantization error.

The magnitude of the quantization error is related to the number of bits used to represent each sample. A larger number of bits allows for more discrete values, resulting in smaller quantization steps and thus a smaller error. Conversely, using fewer bits leads to larger quantization steps and a more significant quantization error. This error manifests as a difference between the original analog value and its digital representation.

Quantization error can introduce noise into the digital signal, affecting its fidelity. In applications like audio or image processing, this can lead to a degradation of the perceived quality. Techniques such as dithering, which involves adding a small amount of random noise before quantization, can be employed to reduce the audibility or visibility of quantization artifacts by spreading the error across a wider range of frequencies. The choice of the number of bits in an ADC is a trade-off between the desired signal fidelity and the required storage space or processing power.